Recommendations
0 to acquire · 1 to dispose
| Ticker | Name | Close | Action | Action Price Where to act — a buy-limit on a pullback or a buy-stop above resistance, not the current price. Sanity-check vs Close: on a held name, an action price stranded above the close means the queued level was set when the stock was higher — stale, awaiting reprice. | Stop The risk-defined exit, set below the entry for a long (~1.2–1.8×ATR). A held long whose Stop sits at/above the Close is already below its stop — a stale box awaiting reprice. | 1wk Model estimate of the likely price at this horizon — a forward projection, not a guarantee, and separate from the trade's Action Price or Stop. Where the ML estimator is enabled these are per-ticker calibrated forecasts (~80% coverage band) that are refreshed periodically and can drift between refreshes; otherwise they're interpolated from the signal's swing target. Confirm against the settled close. | 4wk Model estimate of the likely price at this horizon — a forward projection, not a guarantee, and separate from the trade's Action Price or Stop. Where the ML estimator is enabled these are per-ticker calibrated forecasts (~80% coverage band) that are refreshed periodically and can drift between refreshes; otherwise they're interpolated from the signal's swing target. Confirm against the settled close. | 12wk Model estimate of the likely price at this horizon — a forward projection, not a guarantee, and separate from the trade's Action Price or Stop. Where the ML estimator is enabled these are per-ticker calibrated forecasts (~80% coverage band) that are refreshed periodically and can drift between refreshes; otherwise they're interpolated from the signal's swing target. Confirm against the settled close. | Rationale |
|---|---|---|---|---|---|---|---|---|---|
| CRM | SalesForce Inc | 244.16 | SELL 30% | 244.16 | 232.70 | 251.04 (+2.8%) | 267.08 (+9.4%) | 347.93 (+42.5%) | Salesforce broke its risk line and we are taking a third of the position off. The stock has now closed below our 251.77 stop for a second straight session, this time decisively at 244.16 (0.66 ATR through), and conviction is fading with it, the underlying signal downgraded from BUY to WATCH). We are not exiting: the fresh valuation run puts CRM back on the quality shortlist as a genuine bargain (FAIR_VALUE, F-Score 8, MoS -1.88 against a 393.56 DCF anchor), the 12-week view is still +42.5%, and Dreamforce Investor Day on 09-16 is a real near-term catalyst with Agentforce ARR at 5bn and up 240% year on year. So we shrink the risk and keep the upside. | Operational: fires the numeric exit trigger armed in the 09-09 EOD card (2nd consecutive close < 251.77 OR any close < 245.99 - BOTH legs fired at 244.16); signal-based origination per CLAUDE.md s9 (decisive stop break + deteriorating conviction). L2's second leg (negative 12wk) is ABSENT, which is exactly why this is a 30% trim and not an exit. Valuation amplify DOES NOT apply: the regenerated fairvalue prints MoS_FV -1.88, not <= -3, so no tier bump (the stale 09-02 file would have said -3.37 and bumped this to 50% - the Step 3 regeneration changed the size). Catalyst-aware-hold carve-out for a decisive stop break invoked. Residual re-based to a 1.0-ATR stop at 232.70 (the scan's 238.43 is a 0.50-ATR noise line, rejected per the book's >=1.0-ATR adoption convention). ADR-0020 clear (0% erosion; size now reduced anyway). · news |
| NET | Cloudflare, Inc. | 314.18 | NO ACTION | 314.18 | 292.77 | 325.35 (+3.6%) | 351.42 (+11.9%) | 360.88 (+14.9%) | Book's #1 ranked buy and the strongest chart in the universe - blocked, and not on a close call. Cloudflare is the clearest agentic-AI infrastructure story on the board and it just jumped 10.5% in a session on the OpenAI security tie-up, but it fails the fundamental quality screen outright: F-Score 4, free cash flow of on a $100bn market cap (a 0.31% FCF yield), and a share price roughly ten times the model's fair value anchor. We do not open new positions in names that cannot fund themselves, however good the tape looks. | Adj 65 clears the floor EXACTLY and is still blocked - the gate, not conviction, is the binding constraint. Independent caution: a +10.5% single-session melt-up is the L6 gap-chase shape even though it is not an earnings gap. Ledgered optional:true to price what the gate costs (UT-12 advisory d-iii). Re-arm: Quality_Pass True on a future fairvalue refresh. · news |
| SNDK | Sandisk Corporation | 1764.17 | NO ACTION | 1764.17 | 1714.42 | 1823.87 (+3.4%) | 1963.16 (+11.3%) | 2087.60 (+18.3%) | The only buy candidate in the book that passes the fundamental screen, and it still falls six points short. Sandisk is genuinely high quality here - F-Score 8, 5bn of free cash flow, and the freshly regenerated valuation actually upgrades it to FAIR VALUE and puts it on the model's own bargain shortlist - but the raw signal (69) is drifting rather than building, and this book's own record on FailedBreakdown entries is 2-for-6. With NAND contract pricing decelerating from +70% to +10-15% this quarter, there is no reason to force the entry. | 4th consecutive session declined; conviction trail ->, flat-to-drifting. Re-arm: raw conviction >= 75. NOTE the val_adj -7 is inconsistent with the FAIR_VALUE/bargain label - see systemic finding 2. · news |
| LITE | Lumentum Holdings | 988.98 | BUY-STOP | 1006.49 | 816.44 | 1108.11 (+10.1%) | 1386.58 (+37.8%) | 1205.16 (+19.7%) | Highest conviction print in the book (93) and it finally became slot-eligible today - blocked anyway on fundamentals. Lumentum is the purest AI-optical story we can buy and the revenue growth is real, but it converts almost none of it to cash: of free cash flow against a ~$87bn market cap is a 0.35% yield, F-Score 5, and the price sits roughly nine times the fair-value anchor. Adjusted conviction of 83 is the highest number the book will produce this session and it is still not a reason to buy a business that does not generate cash. | ANTICIPATE count now 2 of last 3 (WATCH -> ANTIC -> ANTIC), so the slot gate is CLEARED for the first time; the quality gate is the sole blocker. Independent support for declining: this book's `anticipate` counterfactual is n=10, median 4wk -8.81% ['saved']. Re-arm: Quality_Pass True. · news |
| AMD | Advanced Micro Devices, Inc | 521.09 | BUY-STOP | 519.46 | 456.07 | 557.50 (+7.3%) | 646.26 (+24.4%) | 517.24 (-0.4%) | The single best live watch item in the book, and the only thing stopping it is the calendar. AMD flipped from AVOID to BUY-ANTICIPATE today on a genuinely strong catalyst - MI450 shipping, data-centre revenue guided to double to $70bn in 2027, and Anthropic adding 2 gigawatts on top of OpenAI and Meta - and unlike every other candidate today it PASSES the quality screen (F-Score 8, 4bn of free cash flow). It just needs one more anticipation print to be queueable. | Anticipate slot requires >=2 of the last 3 snapshots; AMD has 1 (AVOID -> AVOID -> ANTIC). Adj 66 already clears the 55 slot floor. RE-ARM TRIGGER: a 2nd consecutive BUY - ANTICIPATE print at the next EOD makes AMD immediately eligible as a QUEUED-ANTICIPATE buy-stop at the printed Entry_Price, 0.5x normal. Counter-evidence to weigh then: this book's `anticipate` counterfactual is n=10, median 4wk -8.81% ['saved']. · news |
| DELL | Dell Technologies Inc | 535.25 | HOLD | 535.25 | 469.94 | n/a | n/a | n/a | Stop ratcheted 467.56 -> 469.94 on the 2.0x ATR chandelier; the scan's own 443.77 line prints BELOW the existing stop and the ratchet never lowers (4th session running - see systemic finding). Conviction recovering, 12wk 762.73 (+42.5%). NOT armed: Gain_ATR 2.26 vs the 3.0 arm, T12 progress 70.18%. 37.24% of equity and 107% of the book's entire unrealized gain - stated per CLAUDE.md s9, not flagged as a diversification problem. ADR-0020 clear (0% erosion). Zero event-horizon coverage on the largest position, 5th session. |
| MU | Micron Technology, Inc. | 1027.77 | HOLD | 1027.77 | 969.00 | n/a | n/a | n/a | Best-improving holding in the book: conviction and the signal upgraded to AnticipationUp_VCP [STRONG BUY], close +2.75% to a new high for the position (+3.92%, +). Stop held at 969.00 - a 2.0x ATR chandelier computes to 935.88, BELOW the existing line, so the ratchet holds; the scan's 907.83 is lower still. 1.28 ATR of room. No ADD: UT-10 bars pyramiding and the entry sits 0.39 ATR above the close. TSMC August revenue (09-10), Oracle FY27 Q1 (09-10 AMC), PPI 09-10 and CPI 09-11 all read through; FQ4 09-30 still missing from the horizon file. |
| TSM | Taiwan Semiconductor Manufacturing C | 435.36 | HOLD | 435.36 | 419.45 | n/a | n/a | n/a | Second session held, -0.61%. PROFIT_ARMED prints again (T12 progress 98.58%) and is INERT per UT-9 - Gain_ATR is -0.26, an arm-on-loss mislabel on an underwater position, not a winner to bank. Stop held at 419.45 (1.56 ATR of room); the scan's 414.95 is lower and the ratchet does not follow it down. TSMC August monthly revenue prints 09-10 - this position's first real test. UT-7 note: the CSV's Held/Avg_Cost columns are CORRECT this session (TSM Y @ 438.02, MU Y @ 989.00), the first clean session in three. |
| AAPL | Apple | 315.34 | HOLD | 315.34 | 314.62 | n/a | n/a | n/a | *** LIVE RISK: the stop sits 0.72 below the close - 0.09 ATR - and will be taken out by any noise at all. NOT WIDENED: the ratchet never lowers and a stop moved to survive noise stops being a risk line. *** The 09-09 'Surprise and Shine' keynote (iPhone 18 Pro + first foldable, John Ternus's first) passed with the stock essentially flat (-0.28%), so the catalyst-aware hold has now expired and the position is left sitting on its own stop at -3.56%. Conviction improved slightly. PROFIT_ARMED is INERT per UT-9 (Gain_ATR -1.47). Forward view is effectively DEAD: 12wk 320.47 vs 1.02x close = 321.65, i.e. ADR-0012 tick (c) is satisfied - but arming is inert on a losing position, so nothing fires today. ARMED EXIT TRIGGER: any close below 314.62 fires a DISPOSE (30% trim) next session, no further confirmation required. |
| NVDA | NVIDIA Corporation | 223.67 | HOLD | 223.67 | 215.95 | n/a | n/a | n/a | PROFIT_ARMED prints (T12 progress 90.54%, over the 90 arm for the first time) and is INERT per UT-9 - Gain_ATR -0.46, the position is underwater at -1.38%. Conviction, a 2-point drift, nowhere near the 20-point collapse that would be a fire tick. 12wk 247.03 vs 1.02x close 228.14 - forward view alive, so tick (c) does not fire either. Stop held at 215.95, 1.13 ATR of room; the scan's 209.97 is lower. NBIS warrant lock-up (~21M shares) 09-11; Nasdaq-100 quarterly changes 09-11; triple witching 09-18. |
| PLTR | Palantir Technologies Inc. | 169.53 | HOLD | 169.53 | 167.07 | n/a | n/a | n/a | Weakest holding: -3.57% five sessions after entry at 175.81, though conviction has stabilised and the 12wk view is still 223.74 (+32.0%). No ADD - the entry sits 0.91 ATR ABOVE the close (a chase, not a pullback) and this book's PullbackBuy_VWAP cell is 0.400/0.250 at n=6, bias -2. Stop held at 167.07 (0.32 ATR of room, tight); the scan's 168.81 is a 0.09-ATR line and is rejected. Watch item - if it closes below 167.07 with the 12wk still positive it is a trim, not an exit, on the CRM template. |
Outcome & track record
Accuracy at the time of this record.
34 Graded decisions Past recommendations old enough to score against what actually happened (target hit / stopped out / closed). This is the evidence base — the bigger the number, the more history behind the calls. Click to inspect every one. See the decisions →
-0.280 1-week calibration How much of the eventual 4-week move typically shows up in the first week (0 = almost none, 1 = most of it). Use it to read the 1-week estimates: a low number means prices drift toward the target slowly, so don't over-read a soft week-1.
53 Published records How many dated end-of-day verdicts exist for this book — the length of the public track record you can browse in the Archive below.